I think I should have used the provided "words.txt" but I completely forgot about it and I instead downloaded the original "The Time Machine" from Project Gutenberg.
I had to convert some unicode chars to their ascii representation using `unidecode` and the I simply compared the two files byte by byte:
from unidecode import unidecode

orig = unidecode(open('gut.txt', 'rb').read().decode('utf-8'))
m = unidecode(open('ttm.txt', 'rb').read().decode('utf-8'))

f = open('flag.txt', 'wb')

for i in range(len(m)):
a = orig[i]
b = m[i]
if a != b:
print i, a, b
Inside the generated file "flag.txt" you can find the flag: `{FLG:1_kn0w_3v3ryth1ng_4b0ut_t1m3_tr4v3ls}`,